19

THEORY OF MIND AND MORAL

COGNITION
Developmental Changes in Integrating Mental States
and Moral Judgments

Kristin Hansen Lagattuta and Hannah J. Kramer

A central way to predict and explain human behavior is to attend to people’s minds: what they desire,
intend, believe, think, and feel emotionally. Because such mentalizing is a ubiquitous part of every-
day social interactions, a major focus of cognitive development research has been to identify age-
related changes in children’s understanding of the mind from infancy through adulthood, including
reasoning about interrelations between different kinds of mental states (e.g., how thoughts can cause
emotions) and between mind and behavior (e.g., how beliefs influence actions)—what is known as a
theory of mind (Flavell, 2004; Lagattuta et al., 2015; Wellman, 2014). Because humans live in complex
social groups, however, our evaluations and decisions are also guided by rules, norms, obligations,
and permissions. There are certain actions that we should or should not do, have to or do not have
to do, and are permitted or are not permitted to do, as well as opportunities for personal control
or jurisdiction. Research has documented significant developmental changes in these sociomoral
concepts from infancy through adulthood as well (Killen & Smetana, 2015; Nucci, 2014; Tomasello,
2020; Turiel, 2015).
During the past fifteen years, researchers have innovated methods that bridge these two, historically
independent lines of research, establishing the investigation of interrelations between theory of mind
and morality as a robust area of empirical inquiry. In this review, we update our previous chapter from
the second edition of this Handbook (Lagattuta & Weller, 2014). We start with children’s ability to inte-
grate intentions into their moral judgments and then examine how age-related changes in children’s
understanding of beliefs, desires, emotions, and thoughts relate to their emerging moral concepts. We
summarize findings from neuroscience research that has identified substantial overlap between neural
regions recruited for theory of mind and for moral judgment tasks. Finally, we discuss connections
among theory of mind, moral judgment, and behavior. We end with suggestions for future directions.
Although we primarily focus on 3- to 12-year-olds, we also incorporate research with younger and
older age groups.

Using Intentions to Inform Moral Judgments

During the first year of life, infants begin to track others’ intentional, goal-directed actions (Wood-
ward & Sommerville, 2000). By 14 to 18 months, children recognize connections between goals

DOI:10.4324/9781003047247-24 305
 Lagattuta and Kramer

and actions (Fawcett & Markson, 2010), discriminate (on a basic level) between actions done “on
purpose” versus “by accident” (Carpenter et al., 1998), recognize what a person is trying to do even
when they fail (Meltzoff, 1995), and distinguish between adults who are unwilling versus unable to
help (Behne et al., 2005). By 3 years of age, children develop explicit knowledge about the distinc-
tion between intentional and unintentional behaviors, and they can further discriminate intentional
behaviors from desires, mistakes, reflexes, and passive movements (Baird & Astington, 2004). These
early concepts about intention form a cornerstone in the development of folk psychology because
they represent children’s earliest construal of people as agents, who engage in deliberate actions, and
experiencers, who subjectively perceive the world (Wellman, 2014).
Intention is also a core feature of moral judgment, in that intentions are often viewed as criterial
for evaluating the moral status of a person’s actions or character (Alicke & Rose, 2010). Not surpris-
ingly, then, research connecting theory of mind and moral development has focused on children’s
emerging abilities to identify and weigh people’s intentions in their moral judgments. In Piaget’s
(1932) early tests of moral cognition, he presented young children with scenarios in which an actor
either causes minor harm when engaged in a prohibited action (e.g., breaks one dish when trying
to sneak a cookie) or causes more severe harm when trying to be helpful (e.g., breaks three dishes
when helping to set the table). Piaget documented that whereas young children typically judged
the children’s “naughtiness” by the severity of the outcome, children older than 8 or 9 years of age
focused on the child’s motives and intentions and judged the “cookie stealer” as more blameworthy
even though he caused less damage. Piaget argued from these results that children transition from
an objective (focus on the severity of the transgression) to a subjective (focus on the intentions of the
person) view of moral responsibility.
Piaget’s findings sparked numerous investigations into children’s ability to integrate intention, motive,
and outcome information to make moral judgments. Here, intention indicates whether an action was
done purposefully versus accidentally. Motive refers to the actor’s reason for the action (Malle, 1999).
These studies showed that 3- to 5-year-olds can base culpability on intention or motive information
when outcome severity is held constant (e.g., they view intentional negative actions as more blame-
worthy than the same negative outcomes caused by accident; Núñez & Harris, 1998). With increasing
age, children develop more sophisticated insight into relations among intentions, desires, motives, and
beliefs. For example, 7-year-olds, but not younger children, evaluate characters more favorably when
they had good versus bad motives for the same intentional action, even if this behavior results in a nega-
tive outcome (Yuill, 1984).
Recently, there has been a resurgence of empirical inquiry into this outcome to intention shift
(Cushman et al., 2013). Results confirm that from 4 to 8 years of age, children place increasing
emphasis on intentions in their moral evaluations, with the most consistent transition being that chil-
dren view accidental harm less negatively between 6 and 8 years (Killen & Smetana, 2015; Margoni &
Surian, 2017). Parallel age-related patterns emerge in evaluations of fairness. Children younger than
6 to 8 years exhibit a weaker emphasis on the proposer’s intentions to make fair or unfair distribu-
tions (Jaroslawska et al., 2020). Still, even 3- to 4-year-olds can privilege intention over outcome
when shown simplified puppet shows, especially when they receive training and their understand-
ing is measured by their social preference for the accidentally versus intentionally harmful puppet
(Hilton & Kuhlmeier, 2019; see also Woo & Hamlin, this volume, for evidence with infants). Thus,
by minimizing cognitive-processing demands, making intention information highly salient, and/or
measuring intention understanding indirectly (e.g., via preference), it may be possible for younger
children to prioritize intentions over outcomes even in cases of harm.
Because negligence and outcome are often confounded (e.g., although unintentional, harm may
have occurred because the actor was not careful), researchers have tested whether negligence affects
moral judgments. Nobes et al. (2009) found that 3- to 8-year-olds judged intentional careless acts
to be more deserving of punishment than intentional, careful acts. In other variations, Nobes et al.

306
 Theory of Mind and Moral Cognition

(2017) explicitly described the carefulness of the accidental harmer and ensured that children fol-
lowed the actor’s intentions and desires (via control questions). These procedures enabled preschool-
ers to view accidental harm as less punishable than intentional harm, but adults still made sharper
distinctions compared to 7- to 8-year-olds who did so more than 4- to 6-year-olds. Mulvey et al.
(2020) manipulated both the negligence (versus carefulness) of the transgressor and the victim. They
found an age-related increase between 3 and 12 years and between childhood and adulthood in
integrating negligence into evaluations of acceptability, blame, and punishment and in recognizing
situations where both transgressors and victims share in the blame.

Using Moral Judgments (or Outcomes) to Infer Intentions

Not only do children and adults use information about an actor’s intentions to make moral judg-
ments (theory of mind to morality), but the reverse also appears true: children’s and adults’ moral
judgments impact their intuitions about the mental states of the actor (morality to theory of mind).
Knobe (2003) presented adults with scenarios in which a chairman receives information that if he
starts a new program, it will increase his company’s profits, but it will harm (or help) the environ-
ment. The chairman says that he “doesn’t care” about harming (or helping) the environment; he
just wants to maximize profits. He starts the program, and the environment is harmed (or helped).
When adults are asked whether the chairman intentionally harmed or helped the environment, there
is asymmetry in their responses. Harming the environment is viewed as intentionally caused, whereas
helping is not. This is known as the “side-effect effect,” and it has been replicated several times and
in multiple countries and languages (Knobe, 2010). In a variation designed for young children,
Leslie et al. (2006) documented the emergence of this valence asymmetry by age 4: children more
often judged that a character who made their friend feel upset versus happy did so “on purpose”
(60% negative vs. 40% positive outcome). Because this study did not involve older age groups, it is
unknown whether the side-effect effect magnifies between childhood and adulthood. Compared
with the adult findings (Knobe, 2003), it appears as though it does (e.g., 85% negative outcome vs.
23% positive outcome; Knobe, 2003).
Building connections to the broader literature on children’s and adults’ social cognition helps elu-
cidate these valence asymmetries in intention attribution. Children and adults generally believe that
human nature is inherently prosocial, with this assumption possibly emerging early in infancy (Hamlin
et al., 2010). For example, when asked to predict whether a character will expect an unfamiliar per-
son to engage in a positive or negative future action, 4- to 10-year-olds and adults anticipate positive
outcomes at above-chance levels (Lagattuta & Sayfan, 2011). This assumption that most people will
follow social norms increases the signal value of negative behaviors; antisocial actions are viewed as
more revealing of character (Cacioppo & Berntson, 1999). Negative and positive moral obligations
are also not structurally parallel: adults and young children view negative duties (refraining from
harming others) as much more obligatory than positive duties (helping others in need; Kahn, 1992;
Lagattuta, 2018; Weller & Lagattuta, 2013, 2014). Thus, a “not caring” attitude about causing harm
is more unusual, socially deviant, and reprehensible than a “not caring” attitude about a positive side
effect of one’s actions (Alicke & Rose, 2010). More broadly, children (Lagattuta & Wellman, 2002)
and adults (Roese, 1997) more often seek to explain causes of negative versus positive events, and
they more often think about counterfactuals or alternative actions a person could have taken when
outcomes are negative and unexpected (versus positive and expected; Feldman et al., 2016). Thus, it
follows that the trials that elicited the highest attributions of intentionality combined negative out-
comes with social norm violations versus positive outcomes with adherence to social norms.
Developmentally, the signal value of social deviance changes with age. Children 7 years and older
and adults view negative versus positive behaviors as more diagnostic of a person’s character, and
they assume greater consistency in negative versus positive behaviors and traits over time (Boseovski,

307
 Lagattuta and Kramer

2010; Kramer et al., 2021). In contrast, younger children assume more consistency in positive behav-
iors and traits, require more instances of negative behaviors to make character attributions, expect
people’s negative behaviors to improve over time, and endorse goodness as more essential to a person
than badness (Heiphetz, 2019; Kramer et al., 2021). Lagattuta and colleagues (Kramer et al., 2020;
Lagattuta & Sayfan, 2013; Lagattuta et al., 2018; Lagattuta & Kramer, 2021a) found that between the
ages of 4 and 10 and between childhood and adulthood, there are increasing asymmetries in how
children evaluate an agent’s prior antisocial versus prosocial behavior when inferring how the target
of those actions will feel, think, and make decisions when they later see that agent or a similar-look-
ing agent. Participants’ eye movement patterns and verbal judgments indicated that as children grow
older, they more heavily weigh evidence of an actor’s past harmful versus helpful behaviors when
making future forecasts, especially in cases where the most recent past was negative. Combined, these
data help explain why 3-year-olds in the study of Leslie et al. (2006) had difficulty acknowledging
that characters did “not care” about causing harm (i.e., they defaulted to the normative viewpoint of
caring about harm) and why the side-effect effect appears to strengthen as children grow older (i.e.,
the signal value of deviance increases). Although the facts are structurally parallel in the harm and
help scenarios, their meaning in relation to folk psychological beliefs are not.
Indeed, this interpretation that failure to adhere to social norms (versus moral violations specifi-
cally) encourages children and adults to attribute intentions has been supported by empirical evi-
dence. For example, Uttich and Lombrozo (2010) found that the side-effect effect in adults was not
specific to moral infractions (e.g., causing harm); it also arose for violations of social norms (e.g.,
changing the color of “gizmos”). Rakoczy et al. (2015) replicated the side-effect effect for morally
neutral norm violations in 4- to 5-year-olds. Papadopoulos and Hayes (2017) extended this further
to show that norm status mattered for negative, positive, and valence ambiguous side effects, with
Proft et al. (2019) demonstrating the boundary that descriptive norm violations must include a social
conformity component. Taken together, these studies reveal that when an agent chooses to act in
ways that go against moral rules or social norms, this contextual information can provide a powerful
signal to both children and adults that this agent behaved intentionally.

False Belief and Moral Judgment

Because reasoning about agents and outcomes requires considering the minds of those involved (as
highlighted in the prior sections) researchers have explored further connections between theory of
mind and moral judgment. Work on false belief understanding, or children’s knowledge that a person
can believe something that is not true, has dominated theory of mind research since Wimmer and
Perner (1983) reported that children younger than 4 to 5 years of age typically predict that a person
will search for an object in line with reality instead of their false belief. A meta-analysis of more than
five hundred studies revealed that despite variations in characters, content, and questioning between
the ages of 2.5 and 5 years, children transition from viewing the mind and the world as one and the
same (the mind accurately reflects reality) to a new awareness that the mind and world are separate
(the mind can misrepresent reality; Wellman et al., 2001).
Given the centrality of false belief to a child’s understanding of the mind, researchers have mea-
sured beliefs about whether knowledge of a rule is a prerequisite for blame. Samland et al. (2016)
presented 5-year-olds and adults with scenarios where a hedgehog and a bear are both taking pencils
from a box, but there is a rule that hedgehogs are not allowed to use the pencils. In one condition,
the hedgehog is aware of this rule, but in the other condition, the hedgehog is ignorant. Later, a new
bear comes to take a pencil, but the box is empty. Participants determined who was to blame. When
the hedgehog was knowledgeable of the rule, children and adults judged that the hedgehog was more
blameworthy than the bear. More importantly, when the hedgehog was ignorant to the rule, adults
judged that he was no longer blameworthy. In contrast, 5-year-olds in the ignorance condition still

308
 Theory of Mind and Moral Cognition

judged that the hedgehog should be blamed. Thus, young children’s attributions of blame appear
more rule-bound than mental-state-guided (see Kalish & Cornelius, 2007; Wang et al., 2011 for fur-
ther evidence that children younger than 7 years infrequently take into account a person’s knowledge
state when assigning blame and punishment).
Killen and colleagues (2011) extended this inquiry from social conventions to moral rules by
examining 3- to 7-year-olds’ judgments about intention and blame when a protagonist knowingly
versus unknowingly destroys a classmate’s possession (e.g., throws a lunch bag containing her cupcake
in the trash). Compared to younger children, 7-year-olds more often judged that the actor had a
false belief about the bag’s contents (e.g., thought it was trash). Children who attributed a false belief
to the accidental transgressor evaluated him or her as having more positive intentions and being less
blameworthy than a knowledgeable transgressor. More recently, Rizzo et al. (2019) found an age-
related decrease between 3 and 13 years of age in negative evaluations of individuals who do not
intend to make false claims, with children’s understanding of mental states (intention, false belief)
partially mediating the relation between age and moral evaluations.
Together, these data suggest that understanding false belief in morally relevant or normative situa-
tions is more complex than in rule-free contexts. That is, whereas the typical passing age for standard
false belief tasks is about 4 to 5 years of age (Wellman et al., 2001), it is closer to 7 years of age when
the false belief has moral implications, with even more protracted development observed in scenarios
that require coordinating multiple mental states and contextual features. Interestingly, there appear
to be differences in the ways in which 5-year-olds, 7-year-olds, and adults incorporate knowledge
states into their judgments of wrongness, depending upon the type of violation (moral versus con-
ventional). That is, children and adults view moral transgressions (e.g., stealing) as more wrong when
the perpetrator did versus did not know that their actions would lead to a harmful outcome or violate
a rule. In contrast, knowledge states were less important for conventional actions (e.g., dress code;
Giffin & Lombrozo, 2018; Proft & Rakoczy, 2019).
In addition to creating false belief tasks that have morally relevant features, researchers have
also tested connections between children’s false belief understanding and their moral judgments.
Children who score higher on theory-of-mind measures exhibit a greater understanding of the
distinction between moral and conventional rules (Ball et al., 2017), show stronger attention to
how mental states impact people’s moral decisions (Chalik et al., 2014), more negatively evaluate
inequitable resource allocations (Mulvey et al., 2016), and more often intend to act prosocially
(Harari & Weinstock, 2020). On the flip side, children with more advanced theory of mind are
also better skilled at telling and maintaining convincing lies (see Lee & Imuta, 2021 for a meta-
analysis and Evans & Lee, this volume). Although most of these studies have examined inter-
relations between false belief understanding and moral judgments concurrently, those that have
tested for longitudinal relations find mixed results. For example, whereas Smetana et al. (2012)
reported that young children’s moral judgments longitudinally predicted more advanced theory
of mind, Lane et al. (2010) found that early theory of mind (false belief) predicted moral reason-
ing at older ages (see also Sodian et al., 2016). Future research using longitudinal approaches is
needed to provide clarity.

When Desires Conflict With Prohibitive Rules

By the age of 2 to 3 years, children demonstrate explicit awareness of basic, prototypical connections
between desires and emotions: getting what you want feels good and not getting what you want feels
bad (Wellman & Woolley, 1990). This knowledge reflects an important advance in early theory of
mind because it shows appreciation that emotions are not determined just by objective situations but
rather by the person’s mental states. Complications arise, however, when this “desire psychology”
conflicts with rules and norms (i.e., you cannot always do or get what you want). Understanding that

309
 Lagattuta and Kramer

desire-emotion relations can inverse (fulfill desire = feel bad; inhibit desire = feel good) in prohibitive
rule contexts poses a conceptual challenge for young children.

Emotions in Desire-Rule Conflict Situations

Although 4- to 6-year-olds judge that breaking moral rules is wrong, they still frequently infer that
victimizing others to obtain a goal (e.g., shove someone to get a desired swing) feels good, what
is known as the “happy victimizer” effect (see Arsenio et al., 2006). Researchers have traditionally
interpreted these findings to indicate that children under age 7 fail to consider the victim’s negative
emotions when evaluating the emotions of the victimizer. Lagattuta (2005), however, replicated
this “happy rule breaker” age shift in scenarios with no clear victims. Moreover, results also showed
that between 4 and 7 years of age, children increasingly judge that exerting willpower (inhibiting
desires) to abide by rules feels good. Analyses of explanations revealed that whereas 4- to 5-year-olds
overwhelmingly explained emotions in relation to desires (e.g., “Because she didn’t get what she
wanted”), older children and adults more frequently referred to rules (e.g., “Because he did what he
was supposed to do”) or potential future consequences (e.g., “Because if she did pick up her sister,
she might have hurt her”). Thus, the “happy rule breaker” and “sad rule abider” effects in younger
children reflect their difficulty coordinating knowledge about mental states with moral reasoning.
Indeed, this developmental timetable matches the age trajectory for the outcome-to-intention shift
reviewed previously (Cushman et al., 2013), showing that between the ages of 4 and 8 children
increasingly “bridge together a psychological perspective with a sociomoral or deontic perspective”
(Lagattuta, 2005, p. 731).
Additional experimental conditions reveal greater flexibility in young children’s reasoning about
emotions in situations where desires conflict with moral or social norms. For example, 4- to 5-year-
olds more frequently explain emotions in relation to rules or potential future consequences when
they are shown characters in desire-rule conflict situations whose emotions mismatch the status of
desire fulfillment—and they are asked to explain the cause (e.g., explain why a person feels good after
following a rule and not getting what they wanted; Lagattuta, 2005). Moreover, if 4- and 5-year-olds
are explicitly told and shown via pictorial thought bubbles that characters are thinking most about
rules or potential future consequences after making willpower or transgression decisions (perspectives
young children rarely infer spontaneously because they prioritize desires), they can demonstrate a
strong understanding that transgression feels bad and willpower feels good on par with adults (Lagat-
tuta, 2008). Thus, 4- to 5-year-olds can reveal sensitivity to the intersecting influence of rules and
norms when inferring emotions if the cognitive-processing demands are reduced via explicit and
salient primes that pull their focus away from desire fulfillment.

Decisions in Desire-Rule Conflict Situations

Empirical studies have also investigated how children prioritize desires versus rules when predicting
behavior. Kalish and Shiverick (2004) presented 5-year-olds, 8-year-olds, and adults with situations
where a character’s preferences (what he or she likes to do) conflicted with a rule. Across age, par-
ticipants judged that characters will do what they like to do and will feel happy doing what they like
to do, but they should follow the rule. Lagattuta, Nucci, et al. (2010) showed 4- to 8-year-olds sce-
narios in which characters’ desires went against parental rules. For some trials, the rule involved moral
concerns (should not cause harm to others), whereas others involved rules that impinged on the
“personal domain” (e.g., should not wear certain clothes). Although there was an increase between
4 and 7 years of age in predicting that people will follow rules and feel good about doing so for tri-
als featuring moral rules, all age groups rarely predicted “compliance + feel good” for personal rule
trials, especially when parents forbade fulfilling desires essential to identity. Explanation data revealed

310
 Theory of Mind and Moral Cognition

that children emphasized goals, desires, and identity more for the personal (versus moral) conflicts,
whereas they prioritized moral-evaluative concerns (e.g., obligation, permission, rules) more for
the moral (versus personal) conflicts. In a related study, Zhao and Kushnir (2019) found that when
justifying characters’ choices to violate or comply with rules, 5- to 7-year-olds from the U.S. focused
more on desires than did children living in China who more often emphasized evaluations (good/
bad). Combined, these findings highlight the necessity of specifying the rule domain (moral, social
convention, personal) and examining potential cultural variations when investigating how children
integrate theory of mind and morality.
Research on concepts of free will—the inference that a person could have chosen to act other-
wise (Monroe & Malle, 2010)—inform these developmental patterns. Although by age 4 children
recognize that moral and social norms can restrict people’s actions (Chernyak & Kushnir, 2014), 6- to
11-year-olds more often endorse than younger children that people can choose to act against their
strong desires or choose to fulfill their desires even when such actions would violate moral or social
norms (Kushnir et al., 2015). Cross-cultural studies, however, have documented variability in these
judgments, with children growing up in China and Singapore exhibiting a weaker endorsement of
free will compared to children living in North America (Chernyak et al., 2019; Zhao et al., 2021).
Thus, with increasing age (but varying by culture), children better integrate their knowledge that
desires motivate behavior with the awareness that behavioral decisions also depend on the individual’s
autonomous will to choose (see Gopnik & Kushnir, 2014).

Sacrificing Personal Desires to Help Others

A further way to investigate children’s reasoning about desire-rule conflicts is to present them with
prosocial dilemmas in which an actor’s desires are pitted against the needs of another person—
situations in which the actor must sacrifice their own desires to help another person in need. Such
contexts are particularly compelling because improvements in children’s empathic responding and
attention to others’ emotions provide the foundation for prosocial motivation as well as higher shar-
ing, cooperation, and helping behaviors (Eisenberg et al., 2015; Malti, Ongley, et al., 2016; Sierksma
et al., 2014). Multiple studies have shown that children exhibit increasing empathy and other-
oriented perspective-taking across the elementary school years, expecting less self-sacrifice when the
cost of helping is high and more prosocial responses when the cost of helping is low (Eisenberg &
Shell, 1986; Kahn, 1992).
To understand better how characteristics of the situation influence children’s decision and emo-
tion judgments, Weller and Lagattuta (2013, 2014) created prosocial dilemmas in which the charac-
ter’s desire, level of sacrifice, and cost and incentives for helping; the needy individual’s responsibility
for his or her predicament; and the needy child’s level of need were tightly controlled. They found
that 5- to 13-year-olds made different judgments based on need level and that, with increasing age,
children attributed more intensely positive emotions to people who self-sacrificed to help. More
complex developmental findings emerged when children reasoned about not helping others. Results
indicated a U-shaped curve with 5- to 6-year-olds and 11- to 13-year-olds attributing more intensely
positive emotions than 7- to 10-year-olds following decisions to ignore a needy person. Similar cur-
vilinear developmental patterns have been documented in related studies investigating how children
and adolescents coordinate personal desires, the needs of others, and features of the situation (Malti,
Chaparro, et al., 2016; Nucci et al., 2018; Ongley & Malti, 2014), indicating that during middle
childhood and adolescence children can increasingly coordinate multiple contextual nuances and
moral ambiguities in their judgments.
Weller and Lagattuta (2013, 2014) also manipulated the social group of the person in need—this
individual was from the actor’s in-group versus out-group by race (Weller & Lagattuta, 2013) or by
gender (Weller & Lagattuta, 2014). Across age, children rated that actors had a stronger obligation to

311
 Lagattuta and Kramer

help racial in-group versus racial out-group members. Moreover, they judged that individuals would
feel happier helping racial in-group members and happier ignoring the needs of the racial out-group
members. In the case of gender, children exhibited more explicit in-group favoritism: not only did
they more frequently predict that people would feel better helping the gender in-group versus out-
group, but they also more frequently judged that actors would help and not harm someone from their
gender in-group versus out-group (see also Lagattuta, 2018).
These data fit with related studies showing that individuals do not equally apply the same level
of moral status and moral concern to all groups (see also Chalik & Rhodes, this volume). Children
share more with family, friends, and racial in-groups versus with strangers and racial out-groups
(Olson & Spelke, 2008; Renno & Shutts, 2015). Moreover, children and adults judge it more morally
imperative to help and not harm close versus distant others (Nucci et al., 2018), “nice” versus “mean”
people (Hachey & Conry-Murray, 2020; Smetana & Ball, 2018), and in-group versus out-group
members (Rhodes & Chalik, 2013). More generally, beliefs about an individual’s capacity to experi-
ence phenomenological states (e.g., pain, feelings, thoughts) affect ideas about ethical treatment and
moral obligation (Gray et al., 2012; Neldner et al., 2018). Children and adults expect out-group
members to experience less pain and more blunted emotions than in-group members (Dore et al.,
2018), and they attribute fewer mental states and exhibit lower speed and accuracy in mind-reading
out-group versus in-group members (Gönültaş et al., 2020; McLoughlin & Over, 2017). Interven-
tion studies reveal, however, that getting individuals to mentalize, or actively consider the mental
states and emotions of out-group members, can facilitate prosocial behaviors in hypothetical and
real-world situations (McLoughlin & Over, 2019; Todd & Galinsky, 2014).

Theory of Mind and Moral Cognition: Evidence from Cognitive Neuroscience

Research exploring the neurological bases of theory of mind and morality provides additional insight
into their interdependence (see also Decety & Howard, this volume). Functional neuroimaging
(fMRI) studies with children and adults have found striking similarities in the brain regions involved
in theory of mind and moral judgments, although the specific regions can vary by task (Eres et al.,
2018; Moran et al., 2011; Young & Saxe, 2009). There also appears to be a developmental shift
between childhood and adulthood from a more visceral, empathic response to viewing others in
distress (activation of the amygdala and orbitofrontal cortex) to heavier reliance on the prefrontal
cortex (Decety & Cowell, 2018). These data suggest that affective arousal—evaluating something as
good/bad, hostile/benign, or threatening/nurturing, as well as empathic reactions to the distress of
others—may be primary in the integration of theory of mind and morality. As children’s knowledge
about mental states becomes more sophisticated, these connections expand to more thorough con-
siderations of desires, intentions, and beliefs and increasingly involve more frontal areas of the brain.
Mental state networks are particularly activated when viewing acts of intentional harm, assigning
blame for negative outcomes, or reasoning about attempted harm in comparison to thinking about
situations involving neutral intent or outcomes (Decety & Cowell, 2018; Young et al., 2011). This
greater activation in brain areas associated with theory of mind for negative outcomes is intriguing in
light of findings reviewed earlier (e.g., side-effect effect, prioritizing negative versus positive actions
when reasoning about mental states and emotions). That is, converging evidence supports the view
that people weigh mental states more heavily in situations involving actual or potential harm versus
neutral or positive consequences.
More broadly, research utilizing neuroscientific methods has shown that moral judgment requires
the interaction of multiple cognitive processes, including a substantial overlap to neural regions iden-
tified as central to theory of mind and emotion processing (see also Greene, 2003). This is not to say
that the theory-of-mind regions are domain-specific either; many of these regions are also related to
working memory, executive control, episodic memory, prospection, outcome probability, utilitarian

312
 Theory of Mind and Moral Cognition

decisions, and thinking about unexpected outcomes (Knutson & Peterson, 2005; Schacter et al.,
2007). As methods continue to advance, more complex models of the neural bases of theory of mind
and morality will emerge, including their integration with more domain-general cognitive processes.

Theory of Mind, Moral Cognition, and Moral Behavior

A review of interrelations between theory of mind and morality would be incomplete without exam-
ining whether children’s understanding of mental states and moral norms converge to promote moral
action. Children begin to exhibit empathy (Zahn-Waxler et al., 1992) and instrumental helping behav-
iors (Warneken & Tomasello, 2009) starting around 18 months. This timing coincides with their devel-
oping understanding that other people can have mental states and emotions that differ from their own
(Repacholi & Gopnik, 1997), improvements in their physical capabilities, and adult scaffolding and
encouragement (Dahl et al., 2017). From 2 to 5 years of age, children increasingly display guilt and
shame after wrongdoings, show concern about repairing harm, attend more closely to others’ transgres-
sions (Lagattuta & Thompson, 2007), and more often protest against those who laugh at or ignore oth-
ers’ distress (Paulus et al., 2020). Between the ages of 3 and 12 years, children develop more advanced
causal knowledge about interrelations among mental states, emotions, and decisions (Kramer & Lagattuta,
2022; Lagattuta, 2014; Lagattuta et al., 2016; Lagattuta & Kramer, 2021b; Lagattuta et al., 2015).
Several studies have documented links between children’s reasoning about mental states and their
moral actions. For example, Eggum and colleagues (2011) found that 4- to 6-year-olds’ emotional
understanding and theory of mind positively related to their prosocial behavior. Similarly, perfor-
mance on theory of mind measures relates to 4- to 6-year-olds’ allocation of resources (Rizzo &
Killen, 2018). Three- to 9-year-olds’ ratings of how happy they expect to feel sharing positively cor-
relates with their actual resource distributions (Christner et al., 2020). Moreover, children and ado-
lescents who act aggressively more often report than non-aggressive youth that individuals will feel
happier breaking rules (Arsenio et al., 2006; Malti & Krettenauer, 2013; see also Colasante et al., this
volume). A recent meta-analysis confirmed this link between socio-cognitive reasoning and actual
prosocial behavior in 2- to 12-year-olds, but this relation was stronger in individuals older than 6
years (Imuta et al., 2016). Evidence that this association strengthens with age suggests that early in life
individuals’ prosocial decisions (and moral behavior more broadly) may be largely shaped by external
guidance and rules, but as children gain more autonomy, they increasingly leverage their mental state
knowledge to motivate their behavior. Indeed, children are more likely to predict that individuals
will help, share, or follow a rule when this decision feels internally driven versus externally imposed
(Chernyak & Kushnir, 2018; Lagattuta, 2005, 2008; Lagattuta, Nucci, et al., 2010).
Prior literature has been mixed as to whether first-person attributions (reasoning about the self)
or third-person judgments (reasoning about others) more closely correlate with children’s actions
(Arsenio et al., 2006; Gummerum et al., 2010). First-person judgments may be particularly useful
for revealing younger but not necessarily older children’s moral behavior (Malti et al., 2009). This is
because with increasing age, children become more sensitive to how they may be evaluated by others
(Asaba & Gweon, 2018)—their “moral reputation” (Engelmann et al., 2012; Sierksma et al., 2014).
These changes motivate children to present themselves in more socially desirable ways (Lagattuta &
Thompson, 2007), potentially leading to more censored first-person responses. Thus, when testing
more sensitive topics (e.g., moral dilemmas involving race relations), third-person reasoning may
provide a more effective measure (Weller & Lagattuta, 2013, 2014).

Concluding Thoughts and Future Directions

Theory of mind and moral cognition are inherently grounded in how we think about ourselves,
other people, social relationships, and social groups. Thus, consideration of rules and norms may be

313
 Lagattuta and Kramer

inseparable from thinking about mental states. Indeed, as we have highlighted throughout this chap-
ter, moral evaluations inform and are informed by mental state reasoning. How children and adults
construe people’s behaviors in relation to desires, intentions, beliefs, and emotions influences their
moral evaluations, including how they assign blame and allocate praise. Children’s and adults’ moral
cognition guides their inferences about people’s actions, dispositions, and mental states. As children
develop more mature concepts about people, minds, and morality, they increasingly integrate and
coordinate these sometimes conflicting perspectives. We end with some ideas for further exploration.
As reviewed, many studies have examined children’s emphasis on outcomes versus mental states
for assigning blame, with findings showing that between the ages of 4 to 8 years, children shift their
focus away from outcomes and toward the mind. When referring to this developmental change as a
“shift,” however, it is important to acknowledge that outcomes continue to affect moral judgments
across the lifespan. For example, in Lagattuta (2005), all age groups, even adults, judged that indi-
viduals’ primary emotions would align with outcomes. Transgressors (desire fulfilled) feel good and
rule abiders (desire inhibited) feel bad. What changed with age was that older children and adults
recognized that rules and anticipated consequences also influence emotions: they most often inferred
that individuals would feel mixed emotions originating from multiple causes. That is, with increasing
age, children better appreciate the numerous contextual elements—both internal and external—that
shape actions and evaluations (see also Turiel, this volume). Moving forward, eye-tracking may be an
effective tool for expanding inquiry. For example, we have found that biases in children’s and adults’
visual attention to negative versus positive social information relate to their judgments about indi-
viduals’ thoughts, emotions, and decisions (Lagattuta & Kramer, 2021a; Lagattuta & Sayfan, 2013).
Yucel et al. (2020) documented that children’s and adults’ eye-movement patterns and pupil dilation
differed when they viewed moral versus social transgressions (see also Decety et al., 2012; Fiedler &
Glökner, 2015). Thus, combining verbal judgments with eye-tracking may provide more compre-
hensive insight into how children and adults hierarchically weigh and derive meaning from multiple
information sources in morally relevant contexts.
Even with a theory that allows for elements outside highly salient outcomes to matter for moral
evaluations, children and adults still need the cognitive skills to implement these beliefs successfully. For
example, improvements in executive function (especially inhibitory control and working memory) may
help children inhibit a dominant focus on outcomes so that they can incorporate additional factors (e.g.,
desires, intentions, beliefs). Although relations between executive function and theory of mind have
been widely studied (Devine & Hughes, 2014; Lagattuta et al., 2010, 2014, 2016, 2018), contributions
of executive function to moral judgments have been less often empirically tested (Zhao et al., 2021).
Counterfactual reasoning (the ability to think about how outcomes could have been different; Roese,
1997), which also undergoes improvement between 3 and 12 years of age (Beck & Riggs, 2014), may
be another necessary skill for advanced moral cognition (Byrne, 2017). For example, reasoning about
negligence requires imagining whether alternative outcomes could have occurred had individual(s)
acted more carefully. Four- to 8-year-olds who have better counterfactual reasoning or who are cued
to the potential alternative outcomes less often expect rule breakers to feel happy (Gummerum et al.,
2013; Lagattuta, 2008). More work is needed to identify cognitive mechanisms that support children’s
emerging ability to multiply consider intersecting psychological and moral factors.
The previous ideas spark meta-scientific questions about the goals of developmental science in this
research area. One technique has been to substantially reduce cognitive task demands to show the
earliest manifestation of a child’s ability to incorporate mental states into moral evaluations (see Vaish &
Tomasello, this volume; Woo & Hamlin, this volume). This approach is imperative because it eluci-
dates the social patterns infants and toddlers find meaningful and, in doing so, provides informative
data about emerging concepts in mind and morality. A complementary strategy is to amplify cognitive
demands so that tasks more closely simulate everyday experience. Yet even in laboratory paradigms
that assess advanced understanding, experimenters provide participants with explicit details about

314
 Theory of Mind and Moral Cognition

key elements (e.g., the actor’s intention and belief), highlighting their relevance for moral evaluation.
Outside of the lab, individuals need to wade through multiple potentially relevant or irrelevant fac-
tors (e.g., rules, outcomes, situational features, structural constraints, intentions, emotions, desires,
beliefs, past experiences, relationships). Children’s and adults’ emphasis on mental state information
may be weaker in real life compared to hypothetical vignette tasks (Lara et al., 2021). Thus, design-
ing measures that involve a greater number of variables and require more active tracking of mental
states would be theoretically important. Bridging these two approaches, we look forward to greater
emphasis on longitudinal studies that assess links between early and more advanced knowledge.
In closing, previous research has laid important groundwork for future scientific inquiry into con-
nections between theory of mind and morality from a developmental perspective. This increasingly
“hot” topic continues to capture the fascination of developmental psychologists, social psychologists,
neuroscientists, and philosophers. These multidisciplinary empirical approaches make this area of
research uniquely situated for innovating how we think about mind and morality.

References
Alicke, M., & Rose, D. (2010). Culpable control or moral concepts? Behavioral and Brain Sciences, 33, 330–331.
https://doi.org/10.1017/S0140525X10001664
Arsenio, W. F., Gold, J., & Adams, E. (2006). Children’s conceptions and displays of moral emotions. In M. Kil-
len & J. G. Smetana (Eds.), Handbook of moral development (pp. 581–609). Lawrence Erlbaum.
Asaba, M., & Gweon, H. (2018). Look, I can do it! Young children forego opportunities to teach others to dem-
onstrate their competence. In C. Kalish, M. Rau, J. Zhu, & T. Rogers (Eds.), Proceedings of the 40th annual
conference of Cognitive Science Society (pp. 106–111). Cognitive Science Society.
Baird, J. A., & Astington, J. W. (2004). The role of mental state understanding in the development of moral cog-
nition and moral action. New Direction for Child and Adolescent Development, 2004, 37–39. doi: 10.1002/cd.96
Ball, C. L., Smetana, J. G., & Sturge-Apple, M. L. (2017). Following my head and my heart: Integrating pre-
schoolers’ empathy, theory of mind, and moral judgments. Child Development, 88, 597–611. https://doi.
org/10.1111/cdev.12605
Beck, S. R., & Riggs, K. J. (2014). Developing thoughts about what might have been. Child Development Perspec-
tives, 8, 175–179. https://doi.org/10.1111/cdep.12082
Behne, T., Carpenter, M. Call, J., & Tomasello, M. (2005). Unwilling versus unable: Infants’ understanding
of intentional action. Developmental Psychology, 41, 328–337. https://doi.org/10.1037/0012-1649.41.2.328
Boseovski, J. J. (2010). Evidence for “rose-colored glasses”: An examination of the positivity bias in young children’s
personality judgments. Child Development Perspectives, 4, 212–218. https://doi.org/10.1111/j.1750-8606.
2010.00149.x
Byrne, R. M. J. (2017). Counterfactual thinking: From logic to morality. Current Directions in Psychological Science,
26, 314–322. https://doi.org/10.1177/0963721417695617
Cacioppo, J. T., & Berntson, G. G. (1999). The affect system: Architecture and operating characteristics. Current
Directions in Psychological Science, 8, 133–137. https://doi.org/10.1111/1467-8721.00031
Carpenter, M., Akhtar, N., & Tomasello, M. (1998). Fourteen- through 18-month-old infants differentially imi-
tate intentional and accidental actions. Infant Behavior & Development, 21, 315–330. https://doi.org/10.1016/
S0163-6383(98)90009-1
Chalik, L., Rivera, C., & Rhodes, M. (2014). Children’s use of categories and mental states to predict social
behavior. Developmental Psychology, 50, 2360–2367. https://doi.org/10.1037/a0037729
Chernyak, N., Kang, C., & Kushnir, T. (2019). The cultural roots of free will beliefs: How Singaporean and
U.S. children judge and explain possibilities for action in interpersonal contexts. Developmental Psychology, 55,
866–876. https://doi.org/10.1037/dev0000670
Chernyak, N., & Kushnir, T. (2014). The self as a moral agent: Preschoolers behave morally but believe in the
freedom to do otherwise. Journal of Cognition and Development, 15, 453–464. https://doi.org/10.1080/1524
8372.2013.777843
Chernyak, N., & Kushnir, T. (2018). The influence of understanding and having choice on children’s prosocial
behavior. Current Opinion in Psychology, 20, 107–110. https://doi.org/10.1016/j.cogpsyc.2017.07.043
Christner, N., Pletti, C., & Paulus, M. (2020). Emotion understanding and the moral self-concept as motivators
of prosocial behavior in middle childhood. Cognitive Development, 55, 100893. https://doi.org/10.1016/j.
cogdev.2020.100893

315